Hi,l
I have a cheap, Chinese linear power supply and would like to check its output for noise, prior to using it to power a new DAC.
I don't have professional test equipment, neither do I want to invest more than a minimum, as I very rarely need it.
I've been looking at:
iCP12 (1mV) - usbStick (PC USB Oscilloscope, DAQ, Logger, PWM, Analog, IO Board) | eBay
Does anyone have any experience of these and can say if this would be suitable to inspect my 5V (through a simple, voltage-lowering circuit) PSU output?
BR.
--Marius--
I have a cheap, Chinese linear power supply and would like to check its output for noise, prior to using it to power a new DAC.
I don't have professional test equipment, neither do I want to invest more than a minimum, as I very rarely need it.
I've been looking at:
iCP12 (1mV) - usbStick (PC USB Oscilloscope, DAQ, Logger, PWM, Analog, IO Board) | eBay
Does anyone have any experience of these and can say if this would be suitable to inspect my 5V (through a simple, voltage-lowering circuit) PSU output?
BR.
--Marius--
No i don't think that will do it. The resolution and sample rate are very low.
IMHO, you'd be better off investing in a used analog oscilloscope. For most audio related work 20MHz is enough (but more is always better). Dual trace is useful for many tests.
If you can afford a digital storage oscilloscope, all the better.
Most of the inexpensive ($100) USB oscilloscopes aren't very good by all accounts.
IMHO, you'd be better off investing in a used analog oscilloscope. For most audio related work 20MHz is enough (but more is always better). Dual trace is useful for many tests.
If you can afford a digital storage oscilloscope, all the better.
Most of the inexpensive ($100) USB oscilloscopes aren't very good by all accounts.
Line noise is low frequency, so the audible 20kHz band is more than sufficient. The amplitude resolution of a digital oscilloscope is very low, since the goal is to have the system noise to be only less than the thickness of a scope trace. About -40dB is usually all you need to get a crisp trace, and that's generally all you get, since people usually want 100MHz bandwidth at the same time and won't pay for both resolution and bandwidth.
I'm not familiar with what's available now, but basic distortion measuring software that you operate with your computer, using your soundcard or whatever audio input hardware you have, will be far better than any oscilloscope for your use. You want software that can display an FFT of the distortion residual. Using this, you can directly see the magnitude of 60Hz, 120Hz, 180Hz (or is it 50Hz, 100Hz and 150Hz in your country?) separate from any hiss or other noise. Given a generic audio input or a simple soundcard, you should easily get a -100dB noise floor, and better than that using an FFT to display the distortion residual.
I know this type of software has been discussed at length on this site, so do a search and see what software you can come up with. The good part is that it's likely that you won't have to buy any hardware other than some connectors to get the signal into your computer. You'll also have distortion measuring software, which is also generally useful, along with whatever other functions the software provides.
An oscilloscope is great for debugging a faulty analog circuit, but not for measuring low level noise - wrong tool.
I'm not familiar with what's available now, but basic distortion measuring software that you operate with your computer, using your soundcard or whatever audio input hardware you have, will be far better than any oscilloscope for your use. You want software that can display an FFT of the distortion residual. Using this, you can directly see the magnitude of 60Hz, 120Hz, 180Hz (or is it 50Hz, 100Hz and 150Hz in your country?) separate from any hiss or other noise. Given a generic audio input or a simple soundcard, you should easily get a -100dB noise floor, and better than that using an FFT to display the distortion residual.
I know this type of software has been discussed at length on this site, so do a search and see what software you can come up with. The good part is that it's likely that you won't have to buy any hardware other than some connectors to get the signal into your computer. You'll also have distortion measuring software, which is also generally useful, along with whatever other functions the software provides.
An oscilloscope is great for debugging a faulty analog circuit, but not for measuring low level noise - wrong tool.
Checking with load is the only measurement that is valid. If your supply is 317-based, this cheap trick will make it among the best available:
D-Noizator: a magic active noise canceller to retrofit & upgrade any 317-based V.Reg.
D-Noizator: a magic active noise canceller to retrofit & upgrade any 317-based V.Reg.
I would use a dummy load at first so you have a known quantity to work with. Once you're happy with it there, try the actual device, preferably while in use, and see how it performs then.
For the coupling a 10uF electrolytic will do fine (observe polarity), or a .68 or 1uF film cap may cause less distortion at the expense of a small drop at low frequencies
For the coupling a 10uF electrolytic will do fine (observe polarity), or a .68 or 1uF film cap may cause less distortion at the expense of a small drop at low frequencies
Looking at the specs, a 1Khz bandwidth is way too limited to measure any form of PSU noise. One can only observe a limited number of harmonics, but not the garbage that is induced by external influences like RF.
More usable PC-scopes (BW > 20 Mhz) are available from $ 60 upwards.
That's an interesting point! Another project I need to start is a recap of my power amplifier. All the input circuits are coupled using 10uF electrolytics, which I am dubious about. Is that good practice or is it a cost saving excercise on the part of the manufacturer and replacing them with films would be better? regarding polarity for the test scenario, Am I only interested in the '+' conductor ("+" on the cap towards the supply), or should I have one on the 0 V conductor as well ("+" away from the supply)?
This limitation applies to the soundcard also (but less so). On the same token it seems unlikely any reasonably well designed linear power supply is going to pick up a significant amount of RF noise.
There's a lot of discussion elsewhere on capacitor types and distortion ... IMHO the differences are largely irrelevant for the majority of audio applications. As usual with component selection, it's a case of the right tool for the job, diminishing returns and compromise (size, cost, performance, other parameters etc).
I think for the purposes of measuring ripple / noise on a 5v supply rail with a relatively high impedance sound card input, just about any capacitor type you have on hand of an appropriate value will do the job well enough.
- rail can go to sound card ground directly assuming the supply is isolated / floating.
+ supply to + end of cap, - end of cap to sound card input.
The cap will essentially "block" the DC (after initially charging up but 5v for a moment shouldn't hurt anything) and allow you to measure whatever remains.
A good soundcard (192Khz sampling) has less limitation than this USB device, thats correct.
Depends on where the RF originates... If it originates from the circuit you must rely on the suppression characteristics of the supply. And there the design of the PSU and the components used come in.. You're not able to evaluate this behaviour with (USB-)scopes with a limited bandwidth.
Although one could use a modified SDR ($25 + some additional cost) to evaluate 100+ kHz phenomena..
As much as the soundcard should get the job done here, what I like about having a bench oscilloscope on hand, even though it's only an old analog scope, is that it's a known quantity, it's instantaneous, it's robust and it has a good range of scaling right there on a dial.
One thing I notice from my research on many of the cheaper USB oscilloscopes is they don't seem to have a good front end like a benchtop scope has. Perhaps I'm mistaken here or missed something but they all seem limited to low voltages. I'm sure someone can set me straight if I got that wrong.
By putting a scope on a power rail, I can spot 50Hz ripple and any kind of oscillation or serious noise right away. And I would argue that if the ripple and noise is low enough for your requirements, there's no real need to go into FFT analysis.
One thing I notice from my research on many of the cheaper USB oscilloscopes is they don't seem to have a good front end like a benchtop scope has. Perhaps I'm mistaken here or missed something but they all seem limited to low voltages. I'm sure someone can set me straight if I got that wrong.
By putting a scope on a power rail, I can spot 50Hz ripple and any kind of oscillation or serious noise right away. And I would argue that if the ripple and noise is low enough for your requirements, there's no real need to go into FFT analysis.
Hard to say. Perhaps just try it out with REW (its free software), and see if you get any meaningful results from it.
Thanks to all! Very helpful thread! The used analog oscilloscopes on Ebay are sady all too expensive, so the only options for a dedicated unit would be one of the cheap, Chinese jobs. If the USB units are not up to the job, the alternative would have to be a handheld device like this one . But for now, the soundcard route looks like the way to go.
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.